Aerial image segmentation tutorial

This is the source code accompanying the Steep tutorial on semantic segmentation of aerial images:

https://steep-wms.github.io/docs/aerial-image-segmentation/

Contents

data/

Contains an aerial image that can be used as input for the workflow from this tutorial.

The image has been downloaded from https://gds.hessen.de. The data is provided by the Hessische Verwaltung für Bodenmanagement und Geoinformation. It is free and can be used for any purpose.

segmentation-service/

An image segmentation service based on Segment Anything and segment-geospatial.

Usage:

python3 segment.py <input_image> <output_image>

steep/

Configuration files needed to run the image segmentation workflow.

tiling-service/

A tiling service that splits an image into smaller tiles.

Usage:

node tile.js <input_image> <output_directory> <number_of_cols>

<number_of_cols> specifies the number of columns and rows to split the image into. A value of 5 means the image will be split into 5x5 tiles.

workflow/

The image segmentation workflow.